• Journal of the Korea Institute of Military Science and Technology
• /
• v.9 no.3
• /
• pp.132-140
• /
• 2006

In this paper, a new fault detection method for the extended Kalman filter, which uses a N-step predictor, is proposed. The N-step predictor performs the only time propagations for N-step intervals without measurement updates and its output is used as a monitoring signal for the fault detection. A consistency between the extended Kalman filter and the N-step predictor is tested to detect a fault. A test statistic is defined by the difference between the extended Kalman filter and the N-step predictor. The proposed method is applied to strapdown inertial navigation system (SDINS). By computer simulation, it is shown that the proposed method detects a fault effectively.

• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.104-109
• /
• 2002

A new robust Kalman filter is designed for the linear discrete-time system with norm-bounded parametric uncertainties. Sum quadratic constraint, which describes the uncertainties of the system, is converted into an indefinite quadratic form to be minimized in indefinite inner product space. This minimization problem is solved by the new robust Kalman filter. Since the new filter is obtained by simply modifying the conventional Kalman filter, robust filtering scheme can be more readily designed using the proposed method in comparison with the existing robust Kalman filters. A numerical example demonstrates the robustness and the improvement of the proposed filter compared with the existing filters.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.349-351
• /
• 2004

The classical dynamic backpropagation learning algorithm has the problems of learning speed and the determine of learning parameter. The Extend Kalman Filter(EKF) is used effectively for a state estimation method for a non linear dynamic system. This paper presents a learning algorithm using Dual Extended Kalman Filter(DEKF) for Fully Recurrent Neural Network(FRNN). This DEKF learning algorithm gives the minimum variance estimate of the weights and the hidden outputs. The proposed DEKF learning algorithm is applied to the system identification of a nonlinear SISO system and compared with dynamic backpropagation learning algorithm.

• International Journal of Aeronautical and Space Sciences
• /
• v.11 no.2
• /
• pp.80-86
• /
• 2010

We propose a spacecraft attitude estimation algorithm using a federated unscented Kalman filter. For nonlinear spacecraft systems, the unscented Kalman filter provides better performance than the extended Kalman filter. Also, the decentralized scheme in the federated configuration makes a robust system because a sensor fault can be easily detected and isolated by the fault detection and isolation algorithm through a sensitivity factor. Using the proposed algorithm, the spacecraft can continuously perform a given mission despite navigation sensor faults. Numerical simulation is performed to verify the performance of the proposed attitude estimation algorithm.

• Journal of Power Electronics
• /
• v.17 no.1
• /
• pp.190-199
• /
• 2017

In this paper, a predictive direct power control (DPC) method based on a Kalman filter is presented for three-phase pulse-width modulation (PWM) rectifiers to improve the performance of rectifiers with source voltages that are distorted with harmonic components. This method can eliminate the most significant harmonic components of the source voltage using a Kalman filter algorithm. In the process of predicting the future real and reactive power to select an optimal voltage vector in the predictive DPC, the proposed method utilizes source voltages filtered by a Kalman filter, which can mitigate the adverse effects of distorted source voltages on control performance. As a result, the quality of the source currents synthesized using the PWM rectifier is improved, and the total harmonic distortion (THD) values are reduced, even under distorted source voltages.

• Proceedings of the IEEK Conference
• /
• 2009.05a
• /
• pp.187-189
• /
• 2009

Position estimation in indoor is significant problem, because GPS which is usually used for outdoor positioning cannot be utilized to indoor positioning. Sensor network can be a solution for the positioning. Recently, chirp spread spectrum(CSS) specified in IEEE 802.15.4a provides an ability of ranging. Based on the results of the ranging, a position of a CSS node can be calculated by using trilateration. In this case, Kalman filter can be applied to the trilateration because of the measurement noise. In this paper, we apply the unscented Kalman filter for the trilateration. The trilateration can be represented to a nonlinear state space equation, and the unscented Kalman filter is suitable for nonlinear state space equation. Through the experimental results. we show the accuracy of the estimated position.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1679-1683
• /
• 1997

In this paper a linear Kalman filter for calibration of gimballed inertial navigation system(GINS) is designed and its performace is analyzed through the simulation with a real navigation data. Simulation results show that the proposed Kalman filter gives a good performance to calibrate the sensor errors.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.12 no.1
• /
• pp.114-122
• /
• 2009

An adaptive Kalman filter is designed as a post-navigation filter to improve the accuracy of GPS receiver's navigation performance in high dynamic environments. Not only the adaptive Kalman filter reduces the large noise error of navigation data which is obtained by least square method, but also the filter is not degraded as normal Kalman filter in high acceleration movements because the system noise is estimated. Also an initialization structure of the filter is desisted in consideration for irregular output condition of navigation data by least squared method such as reacquisition status in GPS receiver. The filter performance is verified by GPS simulator which has the simulation capability of high velocity and acceleration. Finally, a vehicle test including DGPS is executed to conform the real improvement of that filter performance. This filter can be applied to various data measurement systems to improve accuracy in high dynamic conditions besides GPS receiver.

• Nuclear Engineering and Technology
• /
• v.26 no.2
• /
• pp.205-211
• /
• 1994

The best method is selected among the 3 digital dynamic compensation methods which are developed or applied for the Rhodium self-powered neutron detector. The three digital dynamic compensation methods are the existing Dominant Pol Tustin method of the COLSS(Core Operating Limit Supervisory System), the Direct Inversion method and Kalman Filter method. The Direct Inversion method is an improved method of D. Hoppe and R. Maletti and the Kalman Filter method is developed using the Kalman Filter. Response times of the compensated signals to achieve 90% of a step input are 28.1, 17.2 and 6.5 seconds respectively for the same noise gain telling that the Kalman Filter method is the best amens the 3 methods.

• Journal of Institute of Control, Robotics and Systems
• /
• v.9 no.11
• /
• pp.917-924
• /
• 2003

Recurrent neural networks have been successfully applied to communications channel equalization. Major disadvantages of gradient-based learning algorithms commonly employed to train recurrent neural networks are slow convergence rates and long training sequences required for satisfactory performance. In a high-speed communications system, fast convergence speed and short training symbols are essential. We propose decision feedback equalizers using a recurrent neural network trained with Kalman filtering algorithms. The main features of the proposed recurrent neural equalizers, utilizing extended Kalman filter (EKF) and unscented Kalman filter (UKF), are fast convergence rates and good performance using relatively short training symbols. Experimental results for two time-varying channels are presented to evaluate the performance of the proposed approaches over a conventional recurrent neural equalizer.